Commit Graph

800 Commits

Author SHA1 Message Date
Lincoln Stein
8fc75a71ee integrate correctly into app API and add features
- Create abstract base class InvocationStatsServiceBase
- Store InvocationStatsService in the InvocationServices object
- Collect and report stats on simultaneous graph execution
  independently for each graph id
- Track VRAM usage for each node
- Handle cancellations and other exceptions gracefully
2023-08-02 18:10:52 -04:00
Lincoln Stein
ec48779080 blackify 2023-08-02 14:28:19 -04:00
Lincoln Stein
bc20fe4cb5 Merge branch 'main' into feat/select-vram-in-config 2023-08-02 14:27:17 -04:00
Lincoln Stein
5de42be4a6 reduce VRAM cache default; take max RAM from system 2023-08-02 14:27:13 -04:00
Lincoln Stein
29ac252501 blackify 2023-08-02 09:44:06 -04:00
Lincoln Stein
880727436c fix default vram cache size calculation 2023-08-02 09:43:52 -04:00
Lincoln Stein
8a4e5f73aa reset stats on exception 2023-08-01 19:39:42 -04:00
Lincoln Stein
fd7b842419 add execution stat reporting after each invocation 2023-08-01 17:44:09 -04:00
Lincoln Stein
55d3f04476 additional refactoring 2023-07-31 22:36:11 -04:00
Lincoln Stein
72ebe2ce68 refactor root directory detection to be cleaner 2023-07-31 22:30:06 -04:00
Lincoln Stein
7cd8b2f207 Refactor root detection code 2023-07-31 21:15:44 -04:00
Kent Keirsey
746afcd235
Merge branch 'main' into feat/onnx 2023-07-31 16:56:34 -04:00
Brandon Rising
f5ac73b091 Merge branch 'main' into feat/onnx 2023-07-31 10:58:40 -04:00
psychedelicious
2c07f54b6e
Merge branch 'main' into fix-optional 2023-07-31 16:31:01 +10:00
Lincoln Stein
2164da8592 blackify 2023-07-30 16:25:06 -04:00
Lincoln Stein
4121c261a0 fix missing models when INVOKEAI_ROOT="." 2023-07-30 13:37:18 -04:00
Alexandre Macabies
50e00feceb Add missing Optional on a few nullable fields. 2023-07-30 16:25:12 +02:00
Lincoln Stein
982a568349 blackify pr 2023-07-29 10:47:55 -04:00
Lincoln Stein
9968ff2893 fix relative model paths to be against config.models_path, not root 2023-07-29 10:30:27 -04:00
blessedcoolant
bee678fdd1 fix: Prompt Node using incorrect output type 2023-07-29 17:12:25 +12:00
blessedcoolant
aae1670080 fix: Incorrect Prompt Node output type 2023-07-29 16:04:19 +12:00
blessedcoolant
8a0e1b6cfc feat: Create Prompt Input Node 2023-07-29 15:52:37 +12:00
Brandon Rising
6ca0c38ee3 Merge branch 'main' into feat/onnx 2023-07-28 22:06:28 -04:00
ZachNagengast
2164674b01 Black format 2023-07-28 07:49:29 -07:00
Lincoln Stein
e342ca872f fix to work on non-MPS systems 2023-07-28 10:27:49 -04:00
Brandon Rising
da751da3dd Merge branch 'main' into feat/onnx 2023-07-28 09:59:35 -04:00
Brandon Rising
2b7b3dd4ba Run python black 2023-07-28 09:46:44 -04:00
ZachNagengast
ddaf753f7b Merge branch 'set-timestep-mps-fix' of ssh://github.com/ZachNagengast/InvokeAI into set-timestep-mps-fix 2023-07-27 23:40:44 -07:00
ZachNagengast
e6d14c708c Fix variable name 2023-07-27 23:40:23 -07:00
Millun Atluri
7f81a95b20
Merge branch 'main' into set-timestep-mps-fix 2023-07-28 16:12:07 +10:00
psychedelicious
9affdbbaad chore: black 2023-07-28 11:38:52 +10:00
ZachNagengast
aa1f827271 Fix unet_info location, can have no device prop 2023-07-27 14:47:09 -07:00
ZachNagengast
6edeb4e072 Pass device to set_timestep to avoid float64 error 2023-07-27 12:52:18 -07:00
Lincoln Stein
4a30773d09 Merge branch 'main' into feat/unify-logging 2023-07-27 15:25:56 -04:00
Brandon Rising
1ea9ba84f5 Release session if applying ti or lora 2023-07-27 15:20:38 -04:00
Lincoln Stein
64bd11541a Merge branch 'main' into feat/unify-logging 2023-07-27 15:20:07 -04:00
Lincoln Stein
0d8f9cbe55 resolved conflicts with main 2023-07-27 15:11:25 -04:00
Lincoln Stein
fd75a1dd10 reformat with black 2023-07-27 15:01:00 -04:00
Lincoln Stein
8023a23cec beat uvicorn access log into submission 2023-07-27 12:05:17 -04:00
Lincoln Stein
e4c0102b3c unified uvicorn access log entries too 2023-07-27 11:59:29 -04:00
Lincoln Stein
c4a2808a4b use same logging infrastructure for uvicorn and backend 2023-07-27 11:24:07 -04:00
Brandon Rising
59716938bf Remove TensorRT support at the current time until we validate it works, remove time step recorder 2023-07-27 11:18:50 -04:00
Martin Kristiansen
218b6d0546 Apply black 2023-07-27 10:54:01 -04:00
Brandon Rising
f7bb4c3f05 Remove more files no longer needed in main 2023-07-27 10:49:43 -04:00
Brandon Rising
c16da75ac7 Merge branch 'main' into feat/onnx 2023-07-26 10:42:31 -04:00
Lincoln Stein
4bea846199
Merge branch 'main' into feat/safety-checker-node 2023-07-26 10:04:23 -04:00
Lincoln Stein
b8f43f444a implemented startup sanity checks on core models 2023-07-26 08:26:29 -04:00
Lincoln Stein
3e206d4d6a removed nsfw/watermark from invokeai.yaml 2023-07-26 06:53:35 -04:00
Lincoln Stein
2c5b539d3a esrgan and its models are now nested in app config route 2023-07-26 15:27:04 +10:00
Lincoln Stein
85ad5ef204 refactored code; added watermark and nsfw facilities to app config route 2023-07-26 15:27:04 +10:00
Lincoln Stein
5beb11f4e2 tweaks in response to psychedelicious review of PR 2023-07-26 15:27:04 +10:00
Lincoln Stein
bd43751323 update linear graphs to perform safety checking and watermarking 2023-07-26 15:27:04 +10:00
Lincoln Stein
e32cd794f7 add safetychecker and watermark nodes 2023-07-26 15:26:45 +10:00
psychedelicious
75b0507434 feat(nodes): change denoising start/end min/max to 0/1 2023-07-26 14:46:38 +10:00
psychedelicious
61291ea105 feat: sdxl metadata
- update `CoreMetadata` class & `MetadataAccumulator` with fields for SDXL-specific metadata
- update the linear UI graphs to populate this metadata
2023-07-26 14:46:38 +10:00
psychedelicious
840205496a feat(nodes): fix model load events on sdxl nodes
they need the `context` to be provided to emit socket events
2023-07-26 14:46:38 +10:00
psychedelicious
8e90f9024d feat(ui): remove isRefinerAvailable state, update refiner node
We can derive `isRefinerAvailable` from the query result (eg are there any refiner models installed). This is a piece of server state, so by using the list models response directly, we can avoid needing to manually keep the client in sync with the server.

Created a `useIsRefinerAvailable()` hook to return this boolean wherever it is needed.

Also updated the main models & refiner models endpoints to only return the appropriate models. Now we don't need to filter the data on these endpoints.
2023-07-26 14:46:38 +10:00
Lincoln Stein
2aefa921fe fix "unknown model type" error when rebasing a model with API
- Add command-line model probing script for dev use
- Minor documentation tweak
2023-07-25 08:36:57 -04:00
Lincoln Stein
fc4e104c61 tested on 3.11 and 3.10 2023-07-24 17:13:32 -04:00
blessedcoolant
1969afd038
Merge branch 'main' into feat/fix-soft-locks 2023-07-24 20:12:10 +12:00
blessedcoolant
d6bf6513ef
Merge branch 'main' into fix-types-2 2023-07-24 20:01:48 +12:00
blessedcoolant
e766ddbcf4 fix: Generate random seed using the generator instead of RandomState 2023-07-24 19:38:21 +12:00
psychedelicious
66cdeba8a1 fix(nodes): fix seed modulus operation
This was incorect and resulted in the max seed being one less than intended.
2023-07-24 16:44:32 +10:00
psychedelicious
d5a75eb833 feat: increase seed from int32 to uint32
At some point I typo'd this and set the max seed to signed int32 max. It should be *un*signed int32 max.

This restored the seed range to what it was in v2.3.
2023-07-24 16:34:50 +10:00
Alexandre Macabies
07a90c0198 Fix incorrect use of a singleton list.
This was found through pylance type errors. Go types!
2023-07-23 15:28:05 +02:00
psychedelicious
4b334be7d0 feat(nodes,ui): fix soft locks on session/invocation retrieval
When a queue item is popped for processing, we need to retrieve its session from the DB. Pydantic serializes the graph at this stage.

It's possible for a graph to have been made invalid during the graph preparation stage (e.g. an ancestor node executes, and its output is not valid for its successor node's input field).

When this occurs, the session in the DB will fail validation, but we don't have a chance to find out until it is retrieved and parsed by pydantic.

This logic was previously not wrapped in any exception handling.

Just after retrieving a session, we retrieve the specific invocation to execute from the session. It's possible that this could also have some sort of error, though it should be impossible for it to be a pydantic validation error (that would have been caught during session validation). There was also no exception handling here.

When either of these processes fail, the processor gets soft-locked because the processor's cleanup logic is never run. (I didn't dig deeper into exactly what cleanup is not happening, because the fix is to just handle the exceptions.)

This PR adds exception handling to both the session retrieval and node retrieval and events for each: `session_retrieval_error` and `invocation_retrieval_error`.

These events are caught and displayed in the UI as toasts, along with the type of the python exception (e.g. `Validation Error`). The events are also logged to the browser console.
2023-07-23 21:41:01 +10:00
Brandon Rising
78750042f5 Pass in dim overrides 2023-07-21 12:16:24 -04:00
psychedelicious
594bf6fef1 fix(api,ui): fix canvas saved images have extra transparent regions
- add `crop_visible` param to upload image & set to true only for canvas saves
2023-07-21 07:26:12 -04:00
psychedelicious
896b77cf56 feat(api,db): allow creating an image with a board_id 2023-07-21 07:26:12 -04:00
Lincoln Stein
9370572169 prettify startup messages 2023-07-20 22:45:35 -04:00
Brandon Rising
ce08aa350c Allow controlnet passthrough for now 2023-07-20 14:14:04 -04:00
Brandon Rising
4e90376d11 Allow passing in of precision, use available providers if none provided 2023-07-20 13:15:45 -04:00
Lincoln Stein
c013fe5b5d Merge branch 'main' into release/invokeai-3-0-rc 2023-07-20 12:22:27 -04:00
Sergey Borisov
4a0774b260 Use scale from vae 2023-07-20 18:54:51 +03:00
Sergey Borisov
29a590cced Add sdxl generation preview 2023-07-20 18:45:54 +03:00
Lincoln Stein
7deafa838b merge with main 2023-07-20 11:45:54 -04:00
Lincoln Stein
5134de7cfa
Merge branch 'main' into lstein/logger-route 2023-07-20 11:29:48 -04:00
Lincoln Stein
89a15f78dd collapse all autoimport directories into a single folder 2023-07-20 09:01:49 -04:00
Lincoln Stein
603989dc0d added get_log_level and set_log_level operations to the app route 2023-07-20 08:33:01 -04:00
blessedcoolant
267940a77e Merge branch 'main' into feat/controlnet-resize-mode 2023-07-20 22:24:11 +12:00
blessedcoolant
0795d8764f
Merge branch 'main' into fix/post-model-sync 2023-07-20 20:16:14 +12:00
user1
70fec9ddab Added pixel_perfect_resolution() method to controlnet_utils.py, but not using yet. To be usable this will likely require modification of ControlNet preprocessors 2023-07-20 00:41:49 -07:00
user1
909f538fb5 Switching over to controlnet_utils prepare_control_image(), with added resize_mode. 2023-07-20 00:41:49 -07:00
user1
f2f49bd8d0 Added resize_mode param to ControlNet node 2023-07-20 00:41:49 -07:00
user1
b8e0810ed1 Added revised prepare_control_image() that leverages lvmin high quality resizing 2023-07-20 00:41:49 -07:00
user1
6cb9167a1b Added controlnet_utils.py with code from lvmin for high quality resize, crop+resize, fill+resize 2023-07-20 00:41:49 -07:00
psychedelicious
187cf906fa
ui: enhance intermediates clear, enhance board auto-add (#3851)
* feat(ui): enhance clear intermediates feature

- retrieve the # of intermediates using a new query (just uses list images endpoint w/ limit of 0)
- display the count in the UI
- add types for clearIntermediates mutation
- minor styling and verbiage changes

* feat(ui): remove unused settings option for guides

* feat(ui): use solid badge variant

consistent with the rest of the usage of badges

* feat(ui): update board ctx menu, add board auto-add

- add context menu to system boards - only open is select board. did this so that you dont think its broken when you click it
- add auto-add board. you can right click a user board to enable it for auto-add, or use the gallery settings popover to select it. the invoke button has a tooltip on a short delay to remind you that you have auto-add enabled
- made useBoardName hook, provide it a board id and it gets your the board name
- removed `boardIdToAdTo` state & logic, updated workflows to auto-switch and auto-add on image generation

* fix(ui): clear controlnet when clearing intermediates

* feat: Make Add Board icon a button

* feat(db, api): clear intermediates now clears all of them

* feat(ui): make reset webui text subtext style

* feat(ui): board name change submits on blur

---------

Co-authored-by: blessedcoolant <54517381+blessedcoolant@users.noreply.github.com>
2023-07-20 17:44:22 +12:00
Lincoln Stein
d76bf4444c
Update invokeai/app/api/routers/models.py
Co-authored-by: psychedelicious <4822129+psychedelicious@users.noreply.github.com>
2023-07-19 22:46:49 -04:00
Lincoln Stein
170bbd7da3 change GET to POST method for model synchronization route 2023-07-19 22:16:56 -04:00
Lincoln Stein
f2515d9480
fix v1-finetune.yaml is not in the subpath of "" (#3848)
Co-authored-by: Lincoln Stein <lstein@gmail.com>
2023-07-20 14:13:56 +12:00
Lincoln Stein
4d8f17c69d fix v1-finetune.yaml is not in the subpath of "" 2023-07-19 22:06:55 -04:00
Lincoln Stein
77d9657980 don't write root into invokeai.yaml 2023-07-19 21:12:52 -04:00
Lincoln Stein
5f940bf3b3 default precision to "auto" 2023-07-19 18:23:00 -04:00
psychedelicious
509514f11d feat(api): display warning when port is in use 2023-07-19 13:29:31 -04:00
psychedelicious
c557402dbb feat(api): use next available port
Resolves #3515

@ebr @brandonrising can't imagine this would cause issues but just FYI
2023-07-19 13:29:31 -04:00
Lincoln Stein
2fbc6dc315
Merge branch 'main' into bugfix/ImageToLatentsInvocation_fp32_precision 2023-07-19 12:08:04 -04:00
psychedelicious
0724eb9e0a
feat(ui): another go at gallery (#3791)
* feat(ui): migrate listImages to RTK query using createEntityAdapter

- see comments in `endpoints/images.ts` for explanation of the caching
- so far, only manually updating `all` images when new image is generated. no other manual cache updates are implemented, but will be needed.
- fixed some weirdness with loading state components (like the spinners in gallery)
- added `useThumbnailFallback` for `IAIDndImage`, this displays the tiny webp thumbnail while the full-size images load
- comment out some old thunk related stuff in gallerySlice, which is no longer needed

* feat(ui): add manual cache updates for board changes (wip)

- update RTK Query caches when adding/removing single image to/from board
- work more on migrating all image-related operations to RTK Query

* update AddImagesToBoardContext so that it works when user uses context menu + modal

* handle case where no image is selected

* get assets working for main list and boards - dnd only

* feat(ui): migrate image uploads to RTK Query

- minor refactor of `ImageUploader` and `useImageUploadButton` hooks, simplify some logic
- style filesystem upload overlay to match existing UI
- replace all old `imageUploaded` thunks with `uploadImage` RTK Query calls, update associated logic including canvas related uploads
- simplify `PostUploadAction`s that only need to display user input

* feat(ui): remove `receivedPageOfImages` thunks

* feat(ui): remove `receivedImageUrls` thunk

* feat(ui): finish removing all images thunks

stuff now broken:
- image usage
- delete board images
- on first load, no image selected

* feat(ui): simplify `updateImage` cache manipulation

- we don't actually ever change categories, so we can remove a lot of logic

* feat(ui): simplify canvas autosave

- instead of using a network request to set the canvas generation as not intermediate, we can just do that in the graph

* feat(ui): simplify & handle edge cases in cache updates

* feat(db, api): support `board_id='none'` for `get_many` images queries

This allows us to get all images that are not on a board.

* chore(ui): regen types

* feat(ui): add `All Assets`, `No Board` boards

Restructure boards:
- `all images` is all images
- `all assets` is all assets
- `no board` is all images/assets without a board set
- user boards may have images and assets

Update caching logic
- much simpler without every board having sub-views of images and assets
- update drag and drop operations for all possible interactions

* chore(ui): regen types

* feat(ui): move download to top of context menu

* feat(ui): improve drop overlay styles

* fix(ui): fix image not selected on first load

- listen for first load of all images board, then select the first image

* feat(ui): refactor board deletion

api changes:
- add route to list all image names for a board. this is required to handle board + image deletion. we need to know every image in the board to determine the image usage across the app. this is fetched only when the delete board and images modal is opened so it's as efficient as it can be.
- update the delete board route to respond with a list of deleted `board_images` and `images`, as image names. this is needed to perform accurate clientside state & cache updates after deleting.

db changes:
- remove unused `board_images` service method to get paginated images dtos for a board. this is now done thru the list images endpoint & images service. needs a small logic change on `images.delete_images_on_board`

ui changes:
- simplify the delete board modal - no context, just minor prop drilling. this is feasible for boards only because the components that need to trigger and manipulate the modal are very close together in the tree
- add cache updates for `deleteBoard` & `deleteBoardAndImages` mutations
- the only thing we cannot do directly is on `deleteBoardAndImages`, update the `No Board` board. we'd need to insert image dtos that we may not have loaded. instead, i am just invalidating the tags for that `listImages` cache. so when you `deleteBoardAndImages`, the `No Board` will re-fetch the initial image limit. i think this is more efficient than e.g. fetching all image dtos to insert then inserting them.
- handle image usage for `deleteBoardAndImages`
- update all (i think/hope) the little bits and pieces in the UI to accomodate these changes

* fix(ui): fix board selection logic

* feat(ui): add delete board modal loading state

* fix(ui): use thumbnails for board cover images

* fix(ui): fix race condition with board selection

when selecting a board that doesn't have any images loaded, we need to wait until the images haveloaded before selecting the first image.

this logic is debounced to ~1000ms.

* feat(ui): name 'No Board' correctly, change icon

* fix(ui): do not cache listAllImageNames query

if we cache it, we can end up with stale image usage during deletion.

we could of course manually update the cache as we are doing elsewhere. but because this is a relatively infrequent network request, i'd like to trade increased cache mgmt complexity here for increased resource usage.

* feat(ui): reduce drag preview opacity, remove border

* fix(ui): fix incorrect queryArg used in `deleteImage` and `updateImage` cache updates

* fix(ui): fix doubled open in new tab

* fix(ui): fix new generations not getting added to 'No Board'

* fix(ui): fix board id not changing on new image when autosave enabled

* fix(ui): context menu when selection is 0

need to revise how context menu is triggered later, when we approach multi select

* fix(ui): fix deleting does not update counts for all images and all assets

* fix(ui): fix all assets board name in boards list collapse button

* fix(ui): ensure we never go under 0 for total board count

* fix(ui): fix text overflow on board names

---------

Co-authored-by: Mary Hipp <maryhipp@Marys-MacBook-Air.local>
2023-07-19 12:06:38 -04:00
Martin Kristiansen
07c48b2fd1 Moving detected precision to DEFAULT_PRECISION constant 2023-07-19 11:55:37 -04:00
Martin Kristiansen
fface339ae Same fix for ImageToLatentsInvocation 2023-07-19 11:38:13 -04:00
Martin Kristiansen
2ec9dab595 Changing ImageToLatentsInvocation node to default to detected precision instead of fp16 2023-07-19 11:16:00 -04:00